0-AI-MANIFEST.a2ml

# ⚠️ STOP - CRITICAL READING REQUIRED

**THIS FILE MUST BE READ FIRST BY ALL AI AGENTS**

## WHAT IS THIS?

This is the AI manifest for **i-human**. It declares:
- Canonical file locations (where things MUST be, and nowhere else)
- Critical invariants (rules that must NEVER be violated)
- Repository structure and organization

## CANONICAL LOCATIONS (UNIVERSAL RULE)

### Machine-Readable Metadata: `.machine_readable/` ONLY

These 6 a2ml files MUST exist in `.machine_readable/` directory ONLY:
1. **STATE.a2ml** - Project state, progress, blockers
2. **META.a2ml** - Architecture decisions, governance
3. **ECOSYSTEM.a2ml** - Position in ecosystem, relationships
4. **AGENTIC.a2ml** - AI agent interaction patterns
5. **NEUROSYM.a2ml** - Neurosymbolic integration config
6. **PLAYBOOK.a2ml** - Operational runbook

**CRITICAL:** If ANY of these files exist in the root directory, this is an ERROR.

### Bot Directives: `.machine_readable/bot_directives/` ONLY

Bot-specific instructions for your automated agents. Example roles:
- git-ops-bot - Git operations
- quality-bot - Code quality
- deps-bot - Dependency updates
- docs-bot - Documentation
- integration-bot - Integration
- completion-bot - Task completion

### Contractiles: `.machine_readable/contractiles/` ONLY

Policy enforcement contracts:
- k9 - Operational constraints (Nickel)
- dust - Rollback and recovery semantics
- lust - Future intent and roadmap direction
- must - Required invariants and validations
- trust - Cryptographic verification (Haskell)

### Agent Instructions

- `.claude/CLAUDE.md` - Claude-specific patterns (if exists)
- `0-AI-MANIFEST.a2ml` - THIS FILE (universal entry point)

## CORE INVARIANTS

1. **No state file duplication** - Root must NOT contain STATE.a2ml, META.a2ml, etc.
2. **Single source of truth** - `.machine_readable/` is authoritative
3. **No stale metadata** - If root state files exist, they are OUT OF DATE
4. **License consistency** - All code MPL-2.0 unless platform requires MPL-2.0
5. **Author attribution** - Always "Jonathan D.A. Jewell <j.d.a.jewell@open.ac.uk>"

## REPOSITORY STRUCTURE

**I, Human** (Project Veracity) — Digital trust and CAPTCHA risk audit tool.

Measures your CAPTCHA risk by analyzing browser, behavior, and network signals.
A ReScript/Svelte client collects data; an Elixir/Phoenix backend scores and
stores results in ArangoDB using Nickel rules. LMDB for caching.

```
i-human/
├── 0-AI-MANIFEST.a2ml         # THIS FILE (start here)
├── README.adoc                 # Project overview
├── backend/                    # Elixir/Phoenix API (Bandit server)
│   ├── lib/
│   │   ├── backend/            # Core modules (IP, Score, Cache, Arango, Rules)
│   │   └── backend_web/        # Phoenix controllers and router
│   ├── config/                 # Phoenix + Nickel config
│   └── priv/                   # Nickel rule files (audit_rules.ncl)
├── frontend/                   # SvelteKit + ReScript UI
│   ├── src/
│   │   ├── rescript/           # ReScript modules (Fingerprint, etc.)
│   │   ├── lib/                # Svelte lib (API helpers)
│   │   └── routes/             # SvelteKit pages
│   └── bsconfig.json           # ReScript compiler config
├── src/abi/                    # Idris2 ABI definitions
├── ffi/zig/                    # Zig FFI implementation
├── generated/abi/              # Auto-generated C headers
└── .machine_readable/          # ALL machine-readable content
    ├── STATE.a2ml              # Project state, progress, blockers
    ├── META.a2ml               # Architecture decisions, governance
    ├── ECOSYSTEM.a2ml          # Ecosystem position, relationships
    ├── AGENTIC.a2ml            # AI agent interaction patterns
    ├── NEUROSYM.a2ml           # Neurosymbolic integration config
    ├── PLAYBOOK.a2ml           # Operational runbook
    ├── bot_directives/         # Per-bot rules and constraints
    └── contractiles/           # Policy enforcement contracts (k9, dust, lust, must, trust)
```

## SESSION STARTUP CHECKLIST

✅ Read THIS file (0-AI-MANIFEST.a2ml) first
✅ Understand canonical location: `.machine_readable/` (state, bots, contractiles)
✅ Know the invariants (no state file duplication, etc.)
✅ Check for MCP enforcement (if applicable)
✅ Read `.machine_readable/6a2/STATE.a2ml` for current status
✅ Read `.machine_readable/6a2/AGENTIC.a2ml` for interaction patterns

## LIFECYCLE HOOKS

### on-enter (Session Start)

When starting a new session:

1. Read and acknowledge this manifest
2. Log session start (optional but recommended)
   - Format: `[YYYY-MM-DD HH:MM:SS] Session started: [agent-name]`
   - Location: `.machine_readable/session-log.txt`
3. Read `.machine_readable/6a2/STATE.a2ml`
4. Check for blockers
5. State understanding of canonical locations

### on-exit (Session End)

When ending a session:

1. Update `.machine_readable/6a2/STATE.a2ml` if changes made
2. Log session end (optional but recommended)
   - Format: `[YYYY-MM-DD HH:MM:SS] Session ended: [summary]`
   - Location: `.machine_readable/session-log.txt`
3. Document new blockers
4. Summarize outcomes

## ATTESTATION PROOF

After reading this file, demonstrate understanding by stating:

**"I have read the AI manifest. All machine-readable content (state files, bot directives, contractiles) is located in `.machine_readable/` ONLY, and I will not create duplicate files in the root directory."**

## META

- **Format Version:** 1.0.0
- **Created:** 2026-02-20
- **Maintained By:** hyperpolymath
- **License:** MPL-2.0
- **Protocol:** https://github.com/hyperpolymath/0-ai-gatekeeper-protocol